This short issue serves to mark a reminder to mark a longer issue and series of checkpoints for this task.
Per 9/26's meeting with Prof. Foster, we find that in stream-to-stream based transformations, a need for operational semantics will cleanly be able to model how packets should be processed in order via any given DSL program, with leaves organized in FIFO order.
How this extends to policies which model set-to-stream level transformations is not immediately clear. But we intend to use our type system and arbitration to rule out malformed programs with this ambiguity.
KabirSamsi
commented
3 weeks ago
Per 9/26's meeting with Prof. Foster, we find that in stream-to-stream based transformations, a need for operational semantics will cleanly be able to model how packets should be processed in order via any given DSL program, with leaves organized in FIFO order.
How this extends to policies which model set-to-stream level transformations is not immediately clear. But we intend to use our type system and arbitration to rule out malformed programs with this ambiguity.